Block model to Whittle interface
The block model interface is designed to convert a block model into the required format necessary for Whittle to perform an optimisation, using GEOVIA Whittle optimisation and strategic mine planning package.
After the optimisation is complete, it is also possible to read the results of the optimisation back into a block model, either by creating a new block model and filling it with the appropriate data, or by reading Whittle data back into an existing block model. For example, it is possible to output a Whittle file, optimise it within Whittle, and then read the results back into the original block model. You can then visualise the data using standard block model functions.
Several attributes for specific values must exist in the block model, but after these attributes exist, and they are filled with the correct values, it is simple to output the data as a Whittle file.
When outputting a model, you are prompted to input the attributes and values needed for the calculation, and the block size of a Whittle block. The block model is then traversed one block at a time. If the block currently being traversed is smaller than the Whittle block, then the current block becomes a parcel of the Whittle block. If the current block is larger than the Whittle block (a super block), then the super block is divided into its component Whittle size blocks. The size of the Whittle block is always related to the resolution of the block model.
Note that the blocks in the Whittle model file that are output from this interface are not in sorted numerical order. This is because of the nature of the block model, and does not cause any problem to Whittle. However, if you wish to sort your model file, use Whittle's reblocking functionality, without actually changing the block size, just to sort the model.
Model Preparation
To output a correct Whittle file, correct and useful data must first exist in the block model. If the data in the block model is not correct, the Whittle model will not make sense, or it will be incorrect. Therefore, good model preparation is essential.
Note: To conform to guidelines issued by most professional institutions, it is important that you only include mineralised material which is classified as 'measured' or 'indicated' in the pit optimisation process.
Material in the 'inferred' category should be ignored for the purpose of pit optimisation.
Filling the block model attributes with reasonable data is the most difficult part of the interface. But here are some suggestions:
- Ensure all blocks above the topography have zero density and zero grades.
- Ensure that all blocks below the topography do not have zero density, except where modelling voids.
- If using zone numbers to model different slope regions, ensure that all blocks in the model are assigned a slope region.
- Ensure all blocks in the model are assigned a rock code.
- Ensure that mining cost adjustment factors and processing cost adjustment factors, if used, have reasonable values and that background values are filled with 1, not zero or -99.
Required Attributes
First, there must exist an attribute containing a grade value for every block in the model. These values must exist as a fraction of "units" / tonnes, with a zero grade usually denoting an air or waste block. The metal contents produced are then in "units". If an air or waste block has a grade greater than zero, the grade is just ignored as part of the export routine. For a Whittle output, several grade fields are needed, one for each element to process.
Second, there must exist an attribute with a rock code in it. This attribute can be either a character or an integer attribute. This rock code is used to determine whether the block is ore, waste, or air. If the block is an ore block, the rock code is used to determine the material name to output to the Whittle .mod file. Whittle rock codes have a maximum of four characters, so if the code in the block model is longer than four characters, then the names are truncated. Make sure that all rock codes used in the block model are unique over the first four characters. If the rock code is numeric, the number will be converted to a zero-padded character string. Therefore, any numeric rock code larger than 9999 is also truncated. All blocks must have a rock code value, otherwise they are assumed to be waste. Air blocks must have a rock code value.
The third common field is a specific gravity attribute. This attribute contains the specific gravity for the block. If a block is to become a parcel for the Whittle block and the block is not air, its specific gravity value is used in calculating a weighted average of the specific gravity for the entire Whittle block.
There are several attributes which also must exist for some Whittle products but not others.
Whittle .mod files require a Mining Cost Adjustment Factor (MCAF).
All blocks must contain an MCAF value.
However, if all blocks in the model contain the same MCAF, then no attribute is required, the single value can be entered on the export whittle form.
Because cost of mining applies only to blocks containing rock, MCAF can be zero for air blocks.
See Outputting a Whittle model file for more information on how MCAF is used and calculated in sub-blocked models.
A Processing Cost Adjustment Factor (PCAF) must also exist for a Whittle .mod file.
Again, however, if the PCAF value is the same for all blocks in the model, no attribute is required, the value can be entered on the Export Whittle form.
The cost of processing a block obviously only applies if the block is to be processed, so air and waste blocks do not need a meaningful PCAF value.
See Outputting a Whittle model file for more information on how PCAF is used and calculated in sub-blocked models.
If all the above attributes are correctly filled — you can check this graphically using the block model functions — the output Whittle file should be simple to output for all block models except blocks model with attributes filled with Indicator Kriging.
Indicator Kriged Models
Indicator Kriged models are different in that they do not contain a grade as such. They do contain a cutoff and a percentage. The cutoff is a grade value. The percentage determines how much of the block is contained below that cutoff. A mean value is determined from the cutoff and the previous cutoff, and this value is taken as the grade. Each cutoff will make a new parcel of the Whittle block, with a parcel value being determined by multiplying the mean by the specific gravity by the volume. The volume is calculated as the block volume times the percentage below the cutoff. In this way, all cutoffs become parcels of the block.
This is potentially dangerous, as Whittle has a maximum of 999 parcels per block. If each block contains 10 cutoff levels, then a maximum of 99 sub blocks can contain ore, so for an Indicator Kriged model, the resolution used is critical.